JP2019049463A - Method for controlling shape measuring apparatus - Google Patents

Abstract

To provide a method for controlling a shape measuring apparatus that adjusts acceleration/deceleration change timing of a velocity pattern to a control sampling cycle and prevents a control lag.SOLUTION: Based on a shape of a movement path set based on preobtained shape data of an object to be measured, a velocity pattern for moving along the movement path is generated (ST130). It is determined whether velocity change timing of the velocity pattern is adjusted to a control sampling cycle of a shape measuring apparatus. When the velocity change timing of the velocity pattern is not adjusted to the control sampling cycle, the velocity pattern is corrected such that the velocity change timing of the velocity pattern is adjusted to the control sampling cycle of the shape measuring apparatus (ST200).SELECTED DRAWING: Figure 13

Description

  The present invention relates to a control method of a shape measurement apparatus.

There are known shape measuring apparatuses that measure the shape of a measurement object by moving a probe along the surface of the measurement object (see, for example, Patent Documents 1, 2 and 3). In copy measurement, it is necessary to generate a copy measurement path.
The apparatus described in Patent Document 1 converts a design value (for example, NURBS (Non-Uniform Rational B-Spline: non-uniform rational B-spline) data) based on CAD data or the like into a polynomial curve group of a predetermined order.

To briefly explain this procedure, first, CAD data (for example, NURBS data) including path information is received from an external CAD system or the like, and this CAD data is converted into point cloud data. The data of each point is data in which coordinate values (x, y, z) and normal directions (P, Q, R) are combined (that is, (x, y, z, P, Q, R) is there.)
In the present specification, data of a point group having information of (x, y, z, P, Q, R) will be referred to as contour point data for the following description.

  Next, the coordinate value of each point is offset by a predetermined amount in the normal direction. (The predetermined amount is, specifically, the measurer radius r-the reference insertion amount E0.) The point cloud data determined in this manner will be referred to as offset contour point data.

Then, the offset contour point data is converted into a polynomial curve group of a predetermined order.
Here, a cubic function is used as a polynomial and a PCC curve group (Parametric Cubic Curves) is used.
Based on this PCC curve, a path of measurement is generated. Furthermore, the PCC curve is divided into divided PCC curve groups.

  The velocity curve is calculated from the divided PCC curve group to calculate the movement velocity (movement vector) of the probe. For example, the moving speed (moving vector) of the probe is set based on the curvature of each segment of the divided PCC curve group. The probe is moved based on the movement speed calculated in this manner, and the probe is moved along the surface of the measurement object (passive design value scanning measurement).

  Furthermore, there is also known a method of calculating the indentation correction vector from time to time so as to make the indentation amount of the probe constant, and measuring the scanning while correcting the trajectory (active design value scanning measurement).

The "active design value copying measurement" disclosed in Patent Document 2 (Japanese Patent Application Laid-Open No. 2013-238573) will be briefly introduced. In “active design value scanning measurement”, a combined vector V represented by the following (Expression 1) is used as a vector movement command of the probe.
When the probe moves based on the synthetic vector V, the probe (measuring element) moves along the PCC curve, and the workpiece surface copy measurement with constant indentation amount, that is, "active design value copy measurement" is realized. Be done.

  V = Gf x Vf + Ge x Ve + Gc x Vc (Equation 1)

The meaning of the equation will be briefly described with reference to FIG.
In FIG. 1, there is a PCC curve (that is, a copying path) at a position offset from the design data (contour point data) by a predetermined amount (measuring element radius r-reference indentation amount E0).
Further, in FIG. 1, the actual work is slightly deviated from the design data.

  The vector Vf is a path velocity vector. The path velocity vector Vf has a direction from the interpolation point (i) on the PCC curve toward the next interpolation point (i + 1). The magnitude of the path velocity vector Vf is determined based on, for example, the curvature of the PCC curve at the interpolation point (i) (for example, Patent Document 3).

  The vector Ve is a pressing amount correction vector, and is a vector for causing the pressing amount Ep of the probe to be a predetermined reference pressing amount E0 (for example, 0.3 mm). (The indentation amount correction vector Ve is necessarily parallel to the normal of the work surface.)

The vector Vc is a trajectory correction vector. The trajectory correction vector is parallel to the perpendicular drawn from the probe position to the PCC curve.
Gf, Ge, and Gc are a scanning drive gain, an indentation direction correction gain, and an orbit correction gain, respectively.

The PCC curve is illustrated in FIG.
There is a series of PCC curve L_PCC from point P1 to point P7, and this PCC curve L_PCC is divided into a plurality of segments by point P. (Each segment is also a PCC curve.)
The end point of each segment is the start point of the next segment (PCC curve). The coordinates of the start point of the segment are represented as (K _X0 , K _Y0 , K _Z0 ), and the length of a straight line between the start point and the end point in the PCC curve is D.
In this definition, the coordinates {X (S), Y (S), Z (S)} at any position on the PCC curve are coefficients (K _X3 , K _X2 ...) For representing a cubic curve. It is represented by the following formula using K _Z1 and K _Z0 ).

X (S) = K _{X 3} S ³ + K _{X 2} S ² + K _{X 1} S + K _{X 0
^{Y (S) = K Y3 S}} 3 + K Y2 S 2 + K Y1 S + K Y0
^{_{Z (S) = K Z3 S}} 3 + K Z2 S 2 + K Z1 S + K Z0

If the work is a simple shape such as a plane or a circle with a constant curvature, the scanning measurement path may be a simple shape such as a straight line or a circle, and it is not necessary to divide the PCC curve into many segments.
On the other hand, the PCC curve is divided finely to increase the number of segments in order to perform highly accurate scanning measurement while following the measurement site of the workpiece in a complicated shape and accurately following such a measurement site. There must be. For example, in the case of measuring the contour shape of a curve as shown in FIG. 3, it is divided into segments where the curvature changes as exemplified in FIG. Then, an appropriate velocity pattern is set for each segment (Patent Document 3). That is, a velocity pattern that moves as fast as possible while following the curve accurately is set according to each segment.
FIG. 5 shows an example of the velocity pattern.

JP 2008-241420 A JP, 2013-238573, A JP, 2014-21004, A (patent 6063161)

For example, acceleration / deceleration control of the moving speed may be performed in accordance with the speed pattern obtained as shown in FIG. 5, but in actuality, the following problem has occurred.
The velocity pattern of FIG. 5 is obtained by fitting the velocity pattern and arithmetic processing based on the shape such as the curvature of the PCC curve. However, when actually driving and controlling the shape measuring apparatus, the execution of the instruction is performed in control sampling cycle units.
Please refer to FIG. 6 as an example.
FIG. 6 is a diagram in which the velocity pattern is calibrated on the time axis in control sampling cycle units (T). The velocity pattern of FIG. 6 is an example of the case where the control sampling cycle is perfectly matched by accident.
First, acceleration is performed from the velocity Vs, reaches the constant velocity V, advances at the constant velocity V, and then decelerates to the velocity Ve. If acceleration and deceleration of the speed pattern are performed on the set time, the target point should be reached.

On the other hand, in the velocity pattern of FIG. 7, the control sampling period is not met.
The velocity pattern accelerates from the velocity Vs to reach the constant velocity V, but the end time of acceleration does not match the end of the control sampling period. Therefore, acceleration continues to the end of the control sampling period. Then, the target constant velocity V will be exceeded, and the position of the probe will pass the operational target point. Furthermore, after progressing at a constant speed V, transition is made to a deceleration zone, but the start of deceleration is not at the beginning of the control sampling cycle. Therefore, the constant velocity V continues until the end of the control sampling period. If the deceleration timing is delayed, the final position of the probe passes through the operational target point.
As described above, the control error is accumulated whenever there is a difference between the change timing of the acceleration / deceleration and the control sampling period.

  In the drive control of the conventional shape measurement apparatus, the magnitude of the maximum velocity and the maximum acceleration was not so large, so the control deviation did not become fatal. (Or, the maximum velocity and acceleration were suppressed so that the control deviation did not become too large.) However, in the present where improvement in measurement efficiency is desired, the maximum velocity and acceleration become large. The control deviation becomes an important problem to solve.

  An object of the present invention is to provide a control method of a shape measurement apparatus which adapts the timing of acceleration / deceleration change of a velocity pattern to a control sampling period so that a control deviation does not occur.

The control method of the shape measuring device of the present invention is
A control method of a shape measuring apparatus, which moves a probe along a surface of a measurement object to measure the shape of the measurement object,
Based on the shape of the movement path set based on the shape data of the measurement object obtained in advance, a velocity pattern for moving the movement path is generated,
Determining whether the timing of the speed change of the speed pattern conforms to the control sampling period of the shape measuring device;
If the timing of the velocity change of the velocity pattern does not conform to the control sampling period, the velocity pattern is modified such that the timing of the velocity change of the velocity pattern conforms to the control sampling period of the shape measuring device It is characterized by

In the present invention,
When the initial velocity and the final velocity of the velocity pattern are zero and the velocity pattern is a trapezoidal pattern,
Nm is calculated by rounding up the decimal part of the sum of the control sampling number Na of the acceleration section and the control sampling number Nf of the constant speed section,
Assuming that the total movement distance is S and one control sampling period is T, the corrected maximum velocity Vm which is the velocity of the constant velocity section is calculated by S / (T × Nm),
Furthermore, when the acceleration before correction is represented by α, let the value Nam obtained by integerizing the control sampling number Na in the acceleration zone be ROUNDUP (Na ′, 0) (where Na ′ = Vm / (T × α)),
In accordance with this, it is preferable to correct the velocity pattern so as to conform to the control sampling period, as the control sampling number Nfm = Nm-Nam in the constant velocity section.

In the present invention,
When the initial velocity and the final velocity of the velocity pattern are zero and the velocity pattern is a triangular pattern,
If the total movement distance is S, the acceleration before correction is α, and one control sampling cycle is T,
The control sampling number Nan of the integerized acceleration section is set to the following value.
Nan = Roundup (Na, 0)
However, Na is S = α × (Na × T) ² .

Moreover, the control method of the shape measuring device of the present invention is
A control method of a shape measuring apparatus, which moves a probe along a surface of a measurement object to measure the shape of the measurement object,
Based on the shape of the movement path set based on the shape data of the measurement object obtained in advance, a velocity pattern for moving the movement path is generated,
Determining whether the timing of the speed change of the speed pattern conforms to the control sampling period of the shape measuring device;
If the timing of the velocity change of the velocity pattern does not conform to the control sampling period, the velocity pattern is modified such that the timing of the velocity change of the velocity pattern conforms to the control sampling period of the shape measuring device It is characterized by

In the present invention,
If the velocity pattern is a trapezoidal pattern,
Let Na be the number of control samplings in the acceleration zone before correction, Nd be the number of control samplings in the deceleration zone, Nf be the number of control samplings in the constant velocity zone, and Nt be the total number of control samplings.
The total number of control samplings Ntm after correction is the number Ntm, which is obtained by rounding up after the decimal point of the total number of control samplings Nt before the correction,
The number of times of control sampling in the acceleration zone after correction is the value Nam which rounds up the Na,
The number of control samplings in the decelerating section after correction is a value Ndm which is the rounding up of the above
The control sampling number Nft of the constant velocity section after correction is Ntm− (Nam + Ndm) −2,
Furthermore, it is preferable to insert an adjustment section for one control sampling cycle between each of the acceleration section and the constant speed section and between the constant speed section and the deceleration section.

In the present invention,
When the acceleration in the adjustment section is opposite in sign to the acceleration before correction,
Furthermore, it is preferable to subtract 1 from the Nam and subtract 1 from the Ndm.

In the present invention,
If the velocity pattern is a triangular pattern,
Let Na be the number of control samplings in the acceleration zone before correction, Nd be the number of control samplings in the deceleration zone, and Nt be the total number of control samplings.
The total number of control samplings Ntm after correction is the number Ntm, which is obtained by rounding up after the decimal point of the total number of control samplings Nt before the correction,
The number of times of control sampling in the acceleration zone after correction is set to a value Nam, which is obtained by further subtracting 1 from the value obtained by rounding up the Na,
The number of control samplings in the decelerating section after correction is set to a value Ndm obtained by further subtracting 1 from the value obtained by rounding up the Nd,
The control sampling number Nft of the constant velocity section after correction is Ntm− (Nam + Ndm) −2,
Furthermore, it is preferable to insert an adjustment section for one control sampling cycle between each of the acceleration section and the constant speed section and between the constant speed section and the deceleration section.

Moreover, the control method of the shape measuring device of the present invention is
A control method of a shape measuring apparatus, which moves a probe along a surface of a measurement object to measure the shape of the measurement object,
Based on the shape of the movement path set based on the shape data of the measurement object obtained in advance, a velocity pattern for moving the movement path is generated,
Determining whether the timing of the speed change of the speed pattern conforms to the control sampling period of the shape measuring device;
If the timing of velocity change of the velocity pattern does not conform to the control sampling period, the velocity pattern is modified such that the timing of velocity change of the velocity pattern conforms to the control sampling period of the shape measuring device. ,
For the acceleration zone and the deceleration zone of the velocity pattern after the correction,
In the acceleration section, a tentative S-shaped speed curve is calculated on the assumption that the initial speed Vs monotonically accelerates to reach the corrected maximum speed Vm, and the deceleration section monotonously decelerates from the corrected maximum speed Vm to reach the final speed Ve,
In the movement according to the provisional S-shaped velocity curve, the insufficient distance Da is calculated,
The speed compensation amount ΔV is calculated so as to complement the insufficient distance Da with the number of times of control sampling in the acceleration zone or the deceleration zone,
A value obtained by adding the speed complement amount ΔV to the temporary S-curve speed curve at each control sampling timing is defined as a corrected speed.

It is a figure which illustrates the relationship between design data (outline point data), a PCC curve, and an actual work. 3 illustrates a PCC curve. It is a figure which illustrates a curvilinear outline (PCC curve). It is a figure which shows an example which divided | segmented the PCC curve into the segment. It is a figure which shows an example of a velocity pattern. FIG. 10 is a diagram in which the speed pattern is graduated in units of control sampling cycle (T) on the time axis, and the figure shows an example in the case where the speed pattern and the control sampling cycle are perfectly matched. FIG. 10 is a diagram in which a speed pattern is graduated on the time axis in control sampling cycle units (T), and the figure shows an example in which the speed pattern and the control sampling cycle do not match. It is a figure which shows the whole structure of a shape measurement system. FIG. 2 is a functional block diagram of a motion controller and a host computer. It is a figure which shows the structure of a movement instruction | command production | generation part. It is a figure which shows an example of the trapezoid pattern whose start speed Vs and finish speed Ve are zero. It is a figure which shows an example of the triangular pattern whose start speed Vs and finish speed Ve are zero. It is a whole flowchart showing control operation of a motion controller. It is a flowchart for demonstrating the procedure of the speed pattern correction process by a speed pattern correction part. It is a flowchart for demonstrating the procedure of a speed pattern correction process. It is a figure which shows the speed pattern corrected. It is a flowchart for demonstrating the procedure of a speed pattern correction process. It is a figure which shows the speed pattern corrected. It is a figure which illustrates the trapezoid pattern whose initial velocity Vs and final velocity Ve are not zero. It is a flowchart for demonstrating the procedure of a speed pattern correction process. It is a flowchart for demonstrating the procedure of integer conversion processing. It is a flowchart for demonstrating the procedure of integer conversion processing. It is a figure which shows the speed pattern corrected. It is a figure which illustrates the triangular pattern whose initial velocity Vs and final velocity Ve are not zero. It is a flowchart for demonstrating the procedure of a speed pattern correction process. It is a figure which shows the speed pattern corrected. It is a flowchart for demonstrating S character acceleration-deceleration process. FIG. 7 illustrates a temporary S-shaped velocity curve for a simplified corrected velocity pattern. It is a figure for demonstrating the method to obtain an S-shaped speed curve from a linear acceleration-deceleration pattern by S-curve acceleration-deceleration process. It is a figure for demonstrating the method to obtain an S-shaped speed curve from a linear acceleration-deceleration pattern by S-curve acceleration-deceleration process. It is a figure which illustrates the corrected amount of distance. It is a figure which illustrates a speed correction amount.

Embodiments of the present invention are illustrated and described with reference to reference numerals given to elements in the drawings.
First Embodiment
FIG. 8 is a diagram showing an overall configuration of the shape measurement system 100. As shown in FIG.
The shape measurement system 100 includes a three-dimensional measurement machine 200, a motion controller 300 that controls driving of the three-dimensional measurement machine 200, and a host computer 500 that controls the motion controller 300 and executes necessary data processing.

The three-dimensional measuring machine 200 itself is well known, but is illustrated briefly.
The three-dimensional measuring machine 200 includes a surface plate 210, a moving mechanism 220, and a probe 230.

  The moving mechanism 220 is fixed to a portal Y slider 221 slidably provided in the Y direction on the surface plate 210, an X slider 222 sliding along a beam in the X direction of the Y slider 221, and the X slider 222 And a Z spindle 224 for moving up and down in the Z direction in the Z axis column 223.

  A drive motor (not shown) and an encoder (not shown) are attached to the Y slider 221, the X slider 222 and the Z spindle 224, respectively. The drive control signals from the motion controller 300 drive and control the respective drive motors. The encoder detects the amount of movement of each of the Y slider 221, the X slider 222, and the Z spindle 224, and outputs a detected value to the motion controller 300. A probe 230 is attached to the lower end of the Z spindle 224.

  The probe 230 includes a stylus 231 having a measuring element 232 on the tip end side (−Z axis direction side), and a support portion 233 for supporting the base end side (+ Z axis direction side) of the stylus 231. The tracing stylus 232 is spherical and contacts the measurement object W.

  When an external force is applied to the stylus 231, that is, when the tracing stylus 232 abuts on the object to be measured, the support portion 233 can move the stylus 231 in the X, Y, and Z axial directions within a predetermined range. The stylus 231 is supported so that Furthermore, the support portion 233 includes a probe sensor (not shown) for detecting the position of each of the stylus 231 in the axial direction. The probe sensor outputs the detected value to the motion controller 300.

(Configuration of motion controller 300)
FIG. 9 is a functional block diagram of the motion controller 300 and the host computer 500.
The motion controller 300 includes a measurement command acquisition unit 310, a counter unit 330, a movement command generation unit 340, and a drive control unit 350.

  Measurement command acquisition unit 310 acquires PCC curve data from host computer 500. The counter unit 330 counts the detection signals output from the encoder to measure the displacement amount of each slider, and counts the detection signals output from the probe sensor to measure the displacement amount of the probe 230 (stylus 231). . From the measured displacements of the slider and the probe 230, the coordinate position PP of the tracing stylus 232 (hereinafter, the probe position PP) is obtained. Further, from the displacement of the stylus 231 (detection values (Px, Py, Pz) of the probe sensor) measured by the counter unit 330, the pressing amount (absolute value of the vector Ep) of the tracing stylus 232 can be obtained.

The movement command generation unit 340 calculates the movement path of the probe 230 (measuring element 232) for measuring the surface of the measurement object with the probe 230 (measuring element 232), and calculates the velocity vector along the movement path.
The configuration of the movement command generation unit 340 is shown in FIG.
The movement command generation unit 340 includes a speed pattern planning unit 341 and a vector command generation unit 348.
The velocity pattern planning unit 341 includes a velocity pattern calculation unit 342, a velocity pattern correction unit 343, and an acceleration / deceleration adjustment unit 344.
The operation of each functional unit will be described later with reference to the flowchart.

  The drive control unit 350 drives and controls each slider based on the movement vector calculated by the movement command generation unit 340.

The motion controller 300 is connected to a manual controller 400.
The manual controller 400 has a joystick and various buttons, receives a manual input operation from the user, and sends a user's operation command to the motion controller 300. In this case, the motion controller 300 (drive control unit 350) drives and controls each slider in accordance with a user's operation command.

(Configuration of host computer 500)
The host computer 500 includes a CPU 511 (Central Processing Unit), a memory, and the like, and controls the three-dimensional measuring device 200 via the motion controller 300.
The host computer 500 further includes a storage unit 520 and a shape analysis unit 530. The storage unit 520 stores design data (CAD data, NURBS data, etc.) regarding the shape of the measurement object (workpiece) W, measurement data obtained by measurement, and a measurement control program for controlling the entire measurement operation.

  The shape analysis unit 530 calculates surface shape data of the measurement object based on the measurement data output from the motion controller 300, and performs shape analysis for obtaining an error, distortion or the like of the calculated surface shape data of the measurement object. The shape analysis unit 530 is also responsible for calculation processing such as conversion of design data (CAD data, NURBS data, etc.) into PCC curves.

  The measurement operation of the present embodiment is realized by executing the measurement control program by the CPU 511 (central processing unit).

  An output device (display or printer) and an input device (keyboard or mouse) are connected to the host computer 500 as necessary.

(Control method)
A control method of the shape measuring device according to the present embodiment will be described with reference to a flowchart.
As described above, according to the present invention, the timing of acceleration / deceleration change of the speed pattern is adapted to the "control sampling period" so that the control deviation does not occur. It is desirable to correct the speed pattern (for example, FIG. 7) in which the timing of acceleration / deceleration change does not match the control sampling cycle so that the timing of acceleration / deceleration change matches the control sampling cycle.
As the first embodiment, as shown in FIG. 11 and FIG. 12, consider a velocity pattern in which both the start velocity Vs and the end velocity Ve are zero. (A velocity pattern which is not zero at any one of the start velocity Vs and the end velocity Ve will be described in the second embodiment.)

FIG. 13 is an overall flowchart showing the control operation of the motion controller 300.
The motion controller 300 receives PCC curve data as a measurement command generated by the host computer 500 (ST110). The velocity pattern planning unit 341 generates a velocity pattern for scanning and moving the probe 230 according to the PCC curve data (ST120). The velocity pattern calculation unit 342 generates a velocity pattern that moves as fast as possible while following the PCC curve precisely for each segment (ST130).
Although this process itself is the same as in the prior art, for example, the velocity pattern as shown in FIG. 5 described in the background art is generated, but in the first embodiment, as shown in FIG. 11 or FIG. It is assumed that a velocity pattern in which both and the end velocity Ve are zero is generated. For example, suppose that (a segment of) a PCC curve is a simple figure (e.g., a straight line or a curve of constant curvature).

The velocity pattern in FIG. 11 starts from a state where the starting velocity Vs is zero, accelerates with equal acceleration, reaches the maximum velocity Vmax, moves at a constant velocity for a fixed time at this maximum velocity Vmax, and then decelerates at equal deceleration. The end speed ends at Ve = 0.
At this time, it is assumed that the magnitude of the acceleration and the magnitude of the deceleration are the same. That is, the acceleration time and the deceleration time are the same. As in the example of FIG. 11, a velocity pattern having a section of constant velocity movement at the maximum velocity Vmax will be referred to as a trapezoidal pattern.

Here, the magnitude of acceleration (or deceleration) in the acceleration zone and the deceleration zone is set in accordance with the acceleration resistance of the machine (three-dimensional measuring machine). The acceleration (or deceleration) in the acceleration zone and the deceleration zone may be the same as the acceleration resistance of the machine (three-dimensional measuring machine) in the maximum case. Practically, in consideration of safety and smooth acceleration / deceleration operation (specifically, S-curve acceleration / deceleration processing), acceleration (or deceleration) in the acceleration section and the deceleration section is a machine (three-dimensional measuring machine) It is good to set to half of the acceleration resistance of. In this case, assuming that the magnitude of the acceleration resistance of the machine (three-dimensional measuring machine) is β, the magnitude α of the acceleration (or deceleration) in the acceleration zone and the deceleration zone is α = β / 2.
The maximum velocity Vmax in the constant velocity zone is also set in accordance with the speed tolerance of the machine (three-dimensional measuring machine). The maximum velocity Vmax in the constant velocity section can of course be made the same as the maximum velocity of the machine (three-dimensional measuring machine), or safety factor Q (over 1.0, for example 0 .9 etc.).

  The speed pattern in FIG. 12 starts from a state where the start speed Vs is zero, starts to accelerate at equal acceleration, and starts to decelerate before reaching the maximum speed Vmax, and decelerates at equal deceleration, and ends speed Ve Ends with = 0. Also in this case, it is assumed that the magnitude of the acceleration and the magnitude of the deceleration are the same. That is, the acceleration time and the deceleration time are the same. As in the example of FIG. 12, there is no section of constant velocity movement, and a velocity pattern that shifts to the deceleration section immediately after the acceleration section will be referred to as a triangular pattern.

In the speed pattern of FIG. 11 or FIG. 12, the acceleration / deceleration change timing does not match the control sampling cycle. Therefore, the speed pattern correction unit 343 finely adjusts the speed pattern (FIGS. 11 and 12) to match the control sampling cycle (ST200).
In the first embodiment, the velocity pattern after the correction is also moved at the same acceleration (equal deceleration) as before the correction. That is, it is assumed that the acceleration (or deceleration) is constant in the acceleration zone (or the deceleration zone). In other words, the acceleration (or deceleration) is not switched in the acceleration zone (or the deceleration zone).
Also in the corrected velocity pattern, it is assumed that the magnitude of the acceleration and the magnitude of the deceleration are the same. That is, the acceleration time and the deceleration time are the same in the corrected velocity pattern. (The concept of switching the acceleration and the deceleration will be introduced in the second embodiment.)

FIG. 14 is a flowchart for explaining the procedure of the velocity pattern correction process by the velocity pattern correction unit 343.
Since the processing slightly differs depending on whether the velocity pattern is a trapezoidal pattern or a triangular pattern, cases are divided (ST210).
If a maximum distance that can be moved only by a triangular pattern without using a trapezoidal pattern is represented as St, then St = Vmax ² / α. Therefore, it is determined whether the speed pattern becomes a triangular pattern or a trapezoidal pattern depending on whether the movement distance S between the start position and the target position is larger or smaller than the St. Assuming that the movement path is a straight line as a simple case, let S be the movement distance from the start position Ps (Xs, Ys, Zs) to the target position Pe (Xe, Ye, Ze).

S ² = (Xe-Xs) ² + (Ye-Ys) ² + (Ze-Zs) ²

It is.
In the case of S ≦ St, it is a triangular pattern (ST220: NO), and in the case of S> St, it is a trapezoidal pattern (ST220: YES).

Now, consider the case where the velocity pattern is a trapezoidal pattern (FIG. 11) (ST220: YES).
This will be described in order with reference to the flowchart of FIG.
Now, it is assumed that the control sampling number of the acceleration section is represented by Na, and the control sampling number of the constant velocity section is represented by Nf. Since the acceleration time and the deceleration time are the same, the control sampling number Nd of the deceleration section is also Na. The control sampling period is denoted by T.
Then,
(Na + Nf) = S / (T × Vmax) (Equation 1)
Is true.

Equation (1) is derived from the following relationship.
Vmax = α × Na × T
2 × α × Sa = Vmax ² where Sa is the distance of the acceleration zone.
Sf = Nf × T × Vmax Sf is the distance of the constant velocity section.
S = 2 × Sa + Sf

Make the number of control samplings to be an integer value, and first determine the corrected maximum speed Vm according to it (ST 221)
Here, “Na + Nf” in (Expression 1) is a value corresponding to the number of times of control sampling in which the acceleration zone and the constant velocity zone are combined. Therefore, the decimal part is rounded up to generate an integer value Nm such that Nm = ROUNDUP {(Na + Nf), 0}. Then, the velocity Vm corresponding to the integer value Nm is determined.
Since the decimal part of (Na + Nf) is rounded up to Nm, Vm becomes a smaller value than Vmax.

Nm = S / (T × Vm)
Vm = S / (T × Nm) (Expression 1-1)

  (The reason why (Equation 1-1) is obtained will be understood immediately if, for example, moving the triangle for the distance of the deceleration zone to the acceleration zone in FIG. 11)

  The velocity Vm thus obtained is taken as the velocity in the constant velocity section. This Vm is set as the corrected maximum speed (ST221).

Now, the velocity of the constant velocity zone is the corrected maximum velocity Vm.
Here, under the assumption that the acceleration remains α, the number of accelerations Na ′ for reaching the corrected maximum velocity Vm is calculated backward. Then, Na ′ = Vm / (T × α). Then, this acceleration number Na 'is extended to an integer (ST222).
Nam = ROUNDUP {Na ', 0}

  This integer value Nam is the number of times of acceleration into integers. The same number of decelerations is adopted as Nam.

  Then, since the number of times of acceleration Nam is obtained, the number of times of control sampling in the constant velocity section is determined accordingly. That is, Nfm = Nm-Nam. Thus, the control sampling frequency of the constant velocity section is also obtained as the integer value Nfm (ST223).

Assuming that the number of control samplings in the acceleration zone is Nam and the velocity in the constant velocity zone is the corrected maximum velocity Vm, the acceleration αm to be the corrected maximum velocity Vm with the number of accelerations Nam may be determined (ST224).
αm = Vm / (T × Nam)
The acceleration αm thus obtained is taken as the corrected acceleration.

  If the velocity pattern (FIG. 11) is corrected using the corrected acceleration αm and the corrected maximum velocity Vm, the corrected velocity pattern of FIG. 16 is obtained.

(Correction process when velocity pattern is triangular pattern)
Now, returning to the flowchart of FIG. 14, consider the case where the velocity pattern is a triangular pattern (ST220: NO).
The correction process when the velocity pattern is a triangular pattern will be described with reference to the flowchart of FIG.
In this case, the relationship between the movement distance S and the acceleration (deceleration) α is S = α × (Na × T) ^2, where Na is the number of control samplings in the acceleration zone.
First, the number of accelerations Na obtained by this equation is extended to an integer value (ST 231).
Nan = Roundup (Na, 0)

  The integer value Nan is the number of times of acceleration into integers. (The same number of decelerations is adopted as Nan.) Then, an acceleration α n corresponding to the integer value Nan is obtained (ST 232).

S = α n × (Nan × T) ²
αn = S / (Nan × T) ²

  This acceleration α n is a corrected acceleration. When the velocity pattern is corrected using this corrected acceleration, it becomes as shown in FIG. 18 (ST233). Incidentally, when the intermediate maximum speed corresponding to the corrected acceleration α n is represented by V n, it is V n = α n × N an × T.

  As illustrated in FIG. 16 and FIG. 18, the velocity pattern could be corrected so that the timing of acceleration / deceleration change of the velocity pattern matches the “control sampling period” (ST200). After that, the probe may be moved at a moving speed according to this corrected speed pattern. A composite vector command may be generated by the vector command unit (ST140), and drive control of the probe may be executed (ST150). At this time, since the velocity pattern is corrected to conform to the "control sampling period", there is no control deviation, and a highly accurate scanning measurement operation can be realized.

Second Embodiment
Next, a second embodiment of the present invention will be described.
In the second embodiment, even if one or both of the start velocity Vs and the end velocity Ve of the velocity pattern are not zero, the velocity pattern is corrected to conform to the control sampling period.
As a concept of the second embodiment, the value of acceleration (or deceleration) is not changed as much as possible. The adjustment section is inserted between the switching from the acceleration section to the constant speed section and the switching from the constant speed section to the deceleration section.

  The general flow chart of FIG. 13 and the flow chart of FIG. 14 are basically the same in the first embodiment and the second embodiment. Consider the case where the velocity pattern generated by the velocity pattern calculator 342 is, for example, a trapezoidal pattern as shown in FIG. In the velocity pattern of FIG. 19, the initial velocity Vs and the final velocity Ve are not zero.

The processing operation for adjusting the speed pattern to match the control sampling cycle will be sequentially described with reference to the flowchart of FIG.
The speed pattern correction unit 343 calculates the control sampling number Na of the acceleration segment, the control sampling number Nf of the constant velocity segment, and the control sampling number Nd of the deceleration segment (ST 310).
Na = (Vmax-Vs) / (α × T)
Nd == (Vmax-Ve) / (α × T)
Further, the distance of the constant velocity section is represented by Sf.
Nf = Sf / Vmax

  The distance Sf of the constant velocity zone is obtained by subtracting the distance Sa of the acceleration zone and the distance Sd of the deceleration zone from the total movement distance S.

Sf = S-(Sa + Sd)
2 × α × Sa = Vmax ² −Vs ²
2 × α × Sd = Vmax ² −Ve ²

It is determined whether all of the control sampling number Na of the acceleration section, the control sampling number Nf of the constant speed section and the control sampling number Nd of the deceleration section are integer values (ST 320).
If Na, Nf and Nd are all integer values (or within an acceptable range that can be treated as integer values), correction of the velocity pattern is unnecessary (ST 320: YES). If even one of Na, Nf and Nd is not an integer value (ST320: NO), the next integer conversion processing is performed (ST330).
The integer conversion process (ST330) will be described in order with reference to the flowcharts of FIGS. 21 and 22.

First, the total control sampling number Nt is calculated, and further, an integer value Ntm is calculated by rounding up the total control sampling number Nt (ST 331).
Ntm = Roundup (Nt, 0)
Nt = Na + Nf + Nd

Next, the control sampling number Na of the acceleration section and the control sampling number Nd of the deceleration section are rounded down and made integer (ST 332).
Nam = INT (Na)
Ndm = INT (Nd)

Then, the control sampling number Nfm of the integer speed-converted constant velocity section is determined as follows (ST 333).
Nfm = Ntm-(Nam + Ndm)-2 ... (Equation 2)

Here, in (Expression 2), "2" is subtracted at the end.
Now, FIG. 19 schematically shows the relationship between Nfm, Nam and Ndm.
As can be roughly understood in FIG. 19, when the sum of Nfm, Nam and Ndm is compared with the total control sampling number Ntm converted into an integer, the number is two less. An adjustment section for one control sampling cycle is inserted between the switching from the acceleration section to the constant speed section and the switching from the constant speed section to the decelerating section (ST 334) (see also FIG. 23).

The adjustment section to be inserted at the place where the acceleration section is switched to the constant speed section is taken as a first adjustment section.
An adjustment section to be inserted where the constant velocity section is switched to the deceleration section is taken as a second adjustment section.

Then, the acceleration αc1 and αc2 of the adjustment section are obtained (ST 335)
Now, let the acceleration of the first adjustment section be αc1.
α c1 will be referred to as a first correction acceleration.
The second corrected acceleration of the second adjustment section is taken as αc2.
The movement distance of the acceleration section after integerization is Sam,
The movement distance of the constant velocity section after integerization is Sfm,
Let the movement distance of the deceleration section after integerization be Sdm.

In addition, the movement distance of the first adjustment section is S1 m,
The movement distance of the second adjustment section is S2 m.

Sam = Vs × Nam × T + α × (Nam × T) 2/2
Sfm = Vc × Nfm × T
However, Vc = Vs + α × Nam × T + αc × T
Sdm = Ve × Ndm × T + α × (Ndm × T) 2/2

S1m = (Vs + α × Nam × T) × T + αc1 × T 2/2
S2m = (Ve + α × Ndm × T) × T + αc2 × T 2/2

The total travel distance S is
S = Sam + Sfm + Sdm + S1m + S2m
It should be

Also, let Vm be the maximum velocity of the velocity pattern after correction.
The following relationship is required because the speed change needs to be linked to the whole.
Vm = Vs + α × Nam × T + αc1 × T
Vm = Ve + α × Ndm × T + αc2 × T

  The corrected accelerations (decelerations) αc1 and αc2 of the adjustment section can be inversely calculated from the above relational expressions (ST 335).

If the corrected accelerations αc1 and αc2 have exceeded the maximum acceleration α (ST 336), an error signal is generated and termination processing is performed.
(However, theoretically, (the magnitudes of the corrected accelerations αc1 and αc2 calculated by the above calculation should not exceed the (the magnitude of the accelerations α))

Next, it is checked whether the first correction acceleration αc1 is not negative (ST 337).
If the first correction acceleration αc1 is negative, the speed change in the adjustment section is too large. In this case (ST 337: NO), acceleration is excessive in the acceleration section, so the acceleration section is shortened (ST 338), and recalculation is performed.
That is, it is assumed that Nam = INT (Na) -1.
Similarly, it is assumed that Ndm = INT (Nd) -1.

Although it has been described in ST 337 that the first correction acceleration αc 1 should be a positive value, there is one assumption.
In this embodiment, as exemplified in FIG. 5, a predetermined pattern of acceleration → constant velocity → deceleration, constant velocity,... Is applied as a velocity pattern (Patent 6063161), “deceleration → constant velocity → It does not create the section which decelerates suddenly like "acceleration" etc. Therefore, the fact that the first correction acceleration of the first adjustment section becomes negative means that the vehicle is accelerated excessively in the first acceleration section and unnecessary deceleration is performed in the first adjustment section.

  However, the pattern of "deceleration → constant velocity → acceleration" may of course be accepted as a method of creating the velocity pattern. In this case, it may be checked whether the signs of the acceleration in the first acceleration zone (or the deceleration zone) and the acceleration in the first adjustment zone have been reversed.

  In this way, when the corrected accelerations αc1 and αc2 of the adjustment section are obtained, the integerization processing is successful, so the velocity pattern is corrected (ST340 in FIG. 20).

Returning to the flowchart of FIG. 14, consider the case where the velocity pattern is a triangular pattern (FIG. 24) (ST 220: NO).
The correction method in the case where the velocity pattern is a triangular pattern (FIG. 24) is basically the same as the trapezoidal pattern, but there is one point to be noted, so it will be described below.
See the flowchart of FIG.
First, the speed pattern correction unit 343 calculates the control sampling number Na of the acceleration section and the control sampling number Nd of the deceleration section (ST310A).
Na and Nd can be obtained by solving the following relational expressions.
Here, Vu is the maximum velocity Vu in this triangular velocity pattern, ta is an acceleration time (seconds), and td is a deceleration time (seconds).

Vu ² −Vs ² = 2 × α × S
Vu ² −Vs ² = 2 × α × S
S = Sa + Sd
Sa = Vs × ta + α × ta 2/2
Sd = Ve × td + α × td 2/2
Na = ta / T
Nd = td / T

  If either Na or Nd is not an integer value, integer conversion processing is performed (ST320A: NO).

The total control sampling number Nt is calculated, and further, an integer value Ntm is calculated by rounding up the total control sampling number Nt (ST 331 A).
Ntm = Roundup (Nt, 0)
Nt = Na + Nd

The control sampling number Na of the acceleration section and the control sampling number Nd of the deceleration section are rounded down to an integer, and "1" is further reduced (ST332A). This is a procedure for ensuring that there is room for inserting the adjustment section. This is the difference from the previous trapezoidal pattern (ST332 in FIG. 21).
Nam = INT (Na) -1
Ndm = INT (Nd)-1

The subsequent integerization process is the same as in the trapezoidal pattern, so refer to the subsequent steps in FIGS. 21 and 22.
After inserting the adjustment section (ST334), the corrected accelerations αc1 and αc2 may be obtained (ST335). Then, as illustrated in FIG. 26, a velocity pattern adapted to the control sampling period is obtained.

As the velocity pattern can be corrected so that the timing of acceleration / deceleration change of the velocity pattern conforms to the “control sampling period” as illustrated in FIG. 23 and FIG. 26 (ST 200), after this, the corrected velocity pattern It suffices to move the probe at a moving speed that follows.
The combined vector command may be generated by the vector command generation unit 348 (ST140 in FIG. 13), drive control of the moving mechanism 220 may be executed, and the probe 230 may be moved by copying (ST150). At this time, since the velocity pattern is corrected to conform to the "control sampling period", there is no control deviation, and a highly accurate scanning measurement operation can be realized.
According to the second embodiment, it is possible to correct the velocity pattern so as to conform to the “control sampling period” even for velocity patterns in which the initial velocity or the final velocity is not zero.

Third Embodiment
In the third embodiment, S-curve acceleration / deceleration processing will be described.
In the calculation of the velocity pattern up to now, the acceleration was constant, and hence the velocity was expressed as a linear function with respect to time.
However, in actual machine drive control, smooth control is performed such that acceleration is gradually performed and deceleration is performed gradually. That is, S-curve acceleration / deceleration processing is performed.
After the velocity patterns as shown in FIG. 19 and FIG. 24 are obtained, a method of performing S-curve acceleration / deceleration processing adapted to the velocity patterns has already been proposed by the present applicant (for example, Patent 6050636).

  In the third embodiment, S-curve acceleration / deceleration processing in the case where the number of times of control sampling is performed as in the second embodiment (FIGS. 23 and 26) will be described with reference to the flowchart in FIG. The acceleration / deceleration adjustment processing of the third embodiment is performed by the acceleration / deceleration adjustment unit 344 (FIG. 10).

As a result of correcting FIG. 19 by the processing of the second embodiment, it is assumed that a corrected velocity pattern is obtained as shown in FIG.
Now, first, it is considered that in the acceleration zone, the initial velocity Vs monotonously accelerates to reach the corrected maximum velocity Vm, and in the deceleration zone, the corrected maximum velocity Vm monotonously decelerates to the final velocity Ve (FIG. 28) reference). A tentative S-shaped velocity curve is calculated for this simplified corrected velocity pattern (ST 410). It is already known to make such a simple linear acceleration / deceleration pattern into an S-shaped velocity curve by S-curve acceleration / deceleration processing.

For example, it can be understood by referring to FIGS.
First, the acceleration when accelerating from the initial velocity Vs to the corrected maximum velocity Vm with a linear function in a time of (Nam + 1) × T is represented by αa.
Also, to simplify the equation, let tha be 2 × tha = (Nam + 1) × T.
Then, assuming that the velocity curve is a quadratic curve, the velocity V (t) at time t where 0 <t <tha is
V (t) = Vs + (αa / tha) × t ² ,
(Refer to FIG. 29).
Similarly, the velocity V (t) at time t where tha <t <2tha is
V (t) = Vm- (αa / tha) × ( ² tha-t) ² ,
(See FIG. 30).
In this way, the velocity at each time t can be obtained for the temporary S-curve velocity curve.

Next, distance correction amounts Da and Dd are obtained.
Previously, a tentative S-shaped velocity curve was calculated for the simplified corrected velocity pattern (ST 410).
Therefore, there is a portion that has been discarded (see FIG. 31).
Considering the acceleration zone and the first adjustment zone, the distance correction amount Da is as follows.

Da = {(α-αa) × (Nam × T) ² + (αc 1-αa) × T ² } / 2
As described above, αa represents acceleration when accelerating from the initial velocity Vs to the corrected maximum velocity Vm with a linear function.

Similarly, considering the deceleration zone and the second adjustment zone, the distance correction amount Dd is as follows.
Dd = {-(α-αd) × (Ndm × T) ² + (αc 1-αa) × T ² } / 2
α d represents the deceleration when decelerating from the corrected maximum velocity Vm to the final velocity Ve with a linear function.

Since the correction amounts Da and Dd are insufficient, they may be added to the above-mentioned temporary S-shaped velocity curve (see FIG. 32).
For example, in the acceleration period, per control sampling for the temporary S-curve velocity curve,
The velocity correction amount ΔVa = Da / (Nam × T) may be added.
Similarly, in the deceleration zone, per control sampling for the temporary S-curve velocity curve,
The velocity may be added by the velocity correction amount ΔVd = Dd / (Ndm × T).
(The velocity curve created in this way is no longer a quadratic curve.)

  Since the velocity curve of the pseudo S-curve is obtained in this manner, the probe may be moved at a moving velocity according to this velocity pattern. The velocity pattern is corrected to conform to the “control sampling period”, and the velocity curve of the pseudo S-curve has no control deviation, and has a highly accurate and smooth measurement and measurement of the surface by acceleration and deceleration. realizable.

  The present invention is not limited to the above embodiment, and can be appropriately modified without departing from the scope of the present invention.

100 ... shape measurement system,
200 ... three-dimensional measuring machine,
210: Plate, 220: Movement mechanism, 221: Y slider, 222: X slider, 223: Z axis column, 224: Z spindle,
230 ... probe, 231 ... stylus, 232 ... probe, 233 ... support portion,
300 ... motion controller,
310 ... measurement command acquisition unit, 330 ... counter unit, 340 ... movement command generation unit,
341 ... Speed Pattern Planning Department,
342 ... velocity pattern calculation unit, 343 ... velocity pattern correction unit, 344 ... acceleration / deceleration adjustment unit, 348 ... vector command generation unit,
350: Drive control unit,
400 ... manual controller,
500 host computer 520 storage unit 530 shape analysis unit

Claims (4)

A control method of a shape measuring apparatus, which moves a probe along a surface of a measurement object to measure the shape of the measurement object,
Based on the shape of the movement path set based on the shape data of the measurement object obtained in advance, a velocity pattern for moving the movement path is generated,
Determining whether the timing of the speed change of the speed pattern conforms to the control sampling period of the shape measuring device;
If the timing of the velocity change of the velocity pattern does not conform to the control sampling period, the velocity pattern is modified such that the timing of the velocity change of the velocity pattern conforms to the control sampling period of the shape measuring device And controlling the shape measuring apparatus. In the control method of a shape measuring device according to claim 1,
If the velocity pattern is a trapezoidal pattern,
Let Na be the number of control samplings in the acceleration zone before correction, Nd be the number of control samplings in the deceleration zone, Nf be the number of control samplings in the constant velocity zone, and Nt be the total number of control samplings.
The total number of control samplings Ntm after correction is the number Ntm, which is obtained by rounding up after the decimal point of the total number of control samplings Nt before the correction,
The number of times of control sampling in the acceleration zone after correction is the value Nam which rounds up the Na,
The number of control samplings in the decelerating section after correction is a value Ndm which is the rounding up of the above
The control sampling number Nft of the constant velocity section after correction is Ntm− (Nam + Ndm) −2,
Further, a control method of a shape measuring device, wherein adjustment sections for one control sampling cycle are inserted between the acceleration section and the constant speed section and between the constant speed section and the deceleration section, respectively. In the control method of the shape measuring device according to claim 2,
When the acceleration in the adjustment section is opposite in sign to the acceleration before correction,
Furthermore, the control method of the shape measurement apparatus, wherein 1 is subtracted from Nam and 1 is subtracted from Ndm. In the control method of a shape measuring device according to claim 1,
If the velocity pattern is a triangular pattern,
Let Na be the number of control samplings in the acceleration zone before correction, Nd be the number of control samplings in the deceleration zone, and Nt be the total number of control samplings.
The total number of control samplings Ntm after correction is the number Ntm, which is obtained by rounding up after the decimal point of the total number of control samplings Nt before the correction,
The number of times of control sampling in the acceleration zone after correction is set to a value Nam, which is obtained by further subtracting 1 from the value obtained by rounding up the Na,
The number of control samplings in the decelerating section after correction is set to a value Ndm obtained by further subtracting 1 from the value obtained by rounding up the Nd,
The control sampling number Nft of the constant velocity section after correction is Ntm− (Nam + Ndm) −2,
Further, a control method of a shape measuring device, wherein adjustment sections for one control sampling cycle are inserted between the acceleration section and the constant speed section and between the constant speed section and the deceleration section, respectively.